What is it like to actually interview with an 'AI interviewer' that companies are beginning to introduce in their recruitment processes?
In recent years, AI interviewers have become commonplace, with some companies even recommending their use in their hiring practices.
An AI bot interviewed me for a job. It sucked. - YouTube
I was interviewed by an AI bot for a job | The Verge
https://www.theverge.com/featured-video/892850/i-was-interviewed-by-an-ai-bot-for-a-job
On the computer screen, a female interviewer asks, 'What challenges did you face and how did you deal with them?'
Mr. Field, who is on this side of the PC screen, begins to say, 'I have two questions, so let me start with the first one, which is...'
However, the interviewer interrupted him by replying, 'I understand.'
Field looked confused when his answer was interrupted, and the awkward exchange continued as Field and the interviewer repeatedly said the same thing.
In fact, the interviewer was an AI-generated avatar, and Field was interviewing with an AI interviewer. 'I was being interviewed by an AI, and it was unstable, as I expected,' Field said.
Field acknowledges that AI is everywhere in today's hiring landscape, making it difficult to avoid it. But he points out that having an AI avatar conduct a job interview is a step up. Here, Field investigates why AI interviewers are on the rise and whether they actually work.
According to companies that provide AI interviewer services, prominent companies such as Meta, Netflix, and MasterCard have begun using AI interviewers as part of their initial recruitment screening process.
Companies claim that the benefit of introducing AI interviewers is that they can interact in some way with virtually all applicants, not just a select few.
Prem Kumar, CEO of
Tygren Sloyan, CEO of
In fact, Field decided to try out interviews with AI interviewers provided by three companies: CodeSignal, Humanly, and
While an AI interview is typically like having a one-on-one video call with an AI agent, CodeSignal's AI avatar is a static image, and the person on the other end of the call has their video turned off.
Humanly uses AI avatars that mimic humans almost perfectly as interviewers. Field first interviewed with Humanly, which uses the most human-like avatar.
In the case of Humanly, companies can set the interview time and the tone of the avatar.
'Humanly's AI avatar asked some pretty good questions and interacted well, but the appearance of the avatar itself was approaching the uncanny valley for me,' Field said.
He frankly stated that he had to look away from the AI avatar when answering questions in order to concentrate on his answers.
Humanly claims its AI avatars are unique in that they 'let applicants speak a lot,' Kumar said. 'The average response to an avatar is about 200 words, which is longer than a human screener would have over the phone.'
Next, Field interviewed with CodeSignal's AI interviewer.
An AI interviewer might ask, 'How familiar are you with the technical knowledge of AI and the hardware that powers it? How do you keep up to date on that?'
Field said, 'CodeSignal's interview questions were the most realistic and best of all the ones I tried.'
CodeSignal emphasizes 'what you say' rather than 'how you say it,' so Sloyan encourages applicants to focus on the words they use in their responses.
Field also brought up topics unique to editors, such as 'interactions with public relations people who didn't like my article,' but CodeSigna's AI interviewer was able to converse with him just like a real person.
However, sometimes Field had clearly finished answering a question but the AI interviewer failed to recognize it, leading to awkward moments of silence.
Regarding the introduction of AI interviewers, the developers claim that 'AI interviews will reduce bias and provide a more level playing field,' but this is only true if the AI system itself is free of bias.
In fact, AI systems have been reported to contain a variety of biases, due to the human-created content used to train them, which itself contains various prejudices, including sexism and racism.
Another problem, Field said, is that it's impossible for outsiders to see how these AI systems work to rank candidates. Even if you instruct the AI to reveal its ranking process, the AI systems themselves are black boxes, meaning people can't fully understand the AI's decisions.
Eightfold
Eightfold's AI interviewer didn't use an AI avatar or a photo of his face, but instead just a pulsating sphere on the screen that moved when he spoke. Field said it felt the most robotic and unhuman of the AI interviewers he tested, but he didn't experience any uneasy feelings like the uncanny valley.
Amber Grewal, chief business growth officer at Eightfold, argues that the same questions and the same assessment scores are consistently presented to all candidates during interviews, and that scoring is based on their responses, making it less prone to bias.
Eightfold also uses only the information submitted by candidates with their consent and the data approved by clients under contract for evaluation, and does not collect information such as social media.
He expressed confidence in Eightfold's procedures for eliminating bias.
Humanly and CodeSignal also claim to eliminate bias in their AI systems and conduct regular audits.
Field said he's worried about AI interviewers being the future of job interviews, but if you still have to talk to an AI interviewer, he advised them to speak clearly, use keywords and indicators when responding, and ask one or two questions at the end because of the new year.
In the AI interviewer test, the AI sometimes said, 'This is the first round of screening, so unfortunately we can't reliably obtain those details,' and refused to answer questions.
Related Posts:
